I thought I'd offer some comments on my opinion of influence, for those that want to agree, disagree, discuss, or verbally abuse me about
So, what is influence?
In many ways, I think it's reasonable to say that any stone exerts influence on the board around it, even a 1-1 stone. The problem is, we all know that a) influential play is reasonable, and b) playing on the 1-1 is bad pretty much all of the time, so understanding more about how to use influence seems key to make comments like "3rd line for territory, 4th line for influence" actually turn into a game strategy that makes sense. I've noticed that, in general, the weaker end of the player spectrum seems to really struggle with influence based games, because territory is that much easier to "get".
The winner of a Go game is, obviously the person who earns more points per stone over the course of 200 or so moves. Territorial play is easy to assess, because you can see the points developing in front of your eyes - influential play however, requires both a patience in investing in the position for the future, and just as importantly, a basic understanding of how to convert that investment into points, something that will rarely happen by itself.
To begin this diatribe, I'm going to outline some definitions I tend to use, because I think they help with clarity of purpose when applied to stones on the board. I tend to differentiate quite strongly between thickness and influence, even though they are very similar. This is because the impact the difference has is quite profound on developing a strategic plan. In short, every stone has influence, which is reduced by nearby opposing stones and increased by nearby friendly stones. Thickness is easiest thought of as an accumulation of influence that exceeds a certain rather arbitrary value. The importance of this is related to the stone investment - because a lot more stones are played to create a thick position, the importance of getting as many "points per stone" (see second paragraph) as possible becomes urgent. It is very rare for a thick position facing across an open side of the board not to have a very high level of urgency for both players to manage promptly. If the thick player succeeds in developing effectively from his position, he should stand to make either a large territory that is worth the number of stones played, or alternatively should be able to construct an attack that achieves the same thing. In contrast, the goal of the player without the thickness should be to force the thick player to become overconcentrated. The punishment is not a crazy squeeze or a dastardly attack, more often than not, it's overconcentration: forcing the opponent to accept a small amount of territory for a large number of played stones.
In this second diagram, the top after Black's somewhat premature invasion becomes really important. White has created a genuinely thick position - it's both full of stones and with a noticeable lack of weaknesses. As a result, White's extension here is a fairly vital point for Black to be concerned about, and playing simply to prevent it and contribute to the development of the top left Black stone is unlikely to be terrible. In contrast, if Black is to play elsewhere, White's approach to find a cohesive plan at the top that works with his thickness seems eminently reasonable:
Of course, how to implement that plan isn't always obvious, and this returns to the "points per stone" issue. This actually isn't something there's a simple answer for, as it seems the accumulation of experience will teach it with all the caveats and exceptions much more easily than trying to learn it by rote, but in essence, the last diagram is unwise for White as Black can have an easy time making White overconcentrated. The top right White position is "thick", and has plenty of eyespace potential. As a result, it's not really a target, and building two facing walls like this is a good example of "making territory from thickness" being iffy. Instead, White can choose to approach the top left corner directly, where either result is relatively acceptable:
In this situation, White either gets to develop a moyo from his thickness on a suitably large scale for the number of stones invested, or he gets a corner (better than Black's in the top right) in exchange for the thickness. Black's position at the top, prior to , still has a lot of aji at "x" through "z", whereas White' position at the top right has almost nothing for Black to use as leverage. As such, Black following immediately with has a lot of sense - Whilst it might look like White is building points from thickness by playing at "y", it is likely to do so in sente as the threat of shenanigans against the top Black group is high:
This seems to be risking allowing White to have his cake and eat it, as Black still has some aji issues and not only did White take the corner, but his original thickness has both developed some points and the potential to develop yet further. So, again, Black and White's strategy is focused around the thick positions on the board. Continuing on:
is designed as a pre-emptive protection against "a", as Black's position at the top has solidified to the point where he can play "a" and the corresponding push / squeeze (see hide tag below for the sequence) in the hope of profiting on the left. Alternatively, because of White's thickness in the top right, an alternative could be to work on developing that side of the board, simply because Black, aware of the potential of White creating two directional thickness by bending at "b", will be wanting to approach around "c" as soon as possible. For example:
The thickness starts to quickly create a position which has some considerable White potential, provided Black doesn't get there first. Strategically, it is again apparent that making the most of thickness (or not) happens in just a few moves, and has a large enough game impact to be urgent enough not to neglect.
So, back to the original points on influence and thickness. I see the two normal outcomes of influential stones as being thickness or territory. If you play territorially to secure a few points or give one of your groups safety, you've taken your result immediately. With an influential move, the outcome is generally unclear until further moves have been played in the area. A 4-4 opening stone is a quintessential example of this, where with another 2 stones you have a big corner, but just as often you'll have a big wall of stones and your opponent will have the corner. That doesn't mean to say you wait for your opponent to dictate to you what you're allowed to have. Sometimes the whole board changes in a way where one result becomes good for you, and the other less good, and you dictate the terms before your opponent gets the chance to. For example:
Here, with , Black tries to create a situation where an easy path for White becomes much less clear. Prior to , "a" and the extension, a vaguely possible "b", and invasion at "c" and contact fun with "d" all are ideas White might like to throw around. Suddenly, after , White's options become a lot more limited, and all of the remaining ideas give Black more than he could have previously expected. So, the other aspect to think about with influential moves is what to do to develop them, and I think this is probably the point that most kyu players struggle with. Essentially, there is no way out of learning the variations that are possible for each side, as that determines what aji is in a position. From learning that, a good time to develop the position further is when you can change a local miai situation (very common with 4-4 stones) into a local win-win situation (albeit at the cost of allowing your opponent a move elsewhere). The rest is simply timing
My brain is addled, so I'll stop here, but I figured I'd throw it out there for discussion / debate on the offchance we could get a really good thread going on common examples of influence and development ideas in real games. The ones above are somewhat flawed but just a few of the concepts I thought I could contextually try to get across quickly!
topazg wrote:Territorial play is easy to assess, because you can see the points developing in front of your eyes - influential play however, requires both a patience in investing in the position for the future, and just as importantly, a basic understanding of how to convert that investment into points, something that will rarely happen by itself.
I probably shouldn't be posting a response because it might be beyond my level. I also just skimmed through the explanation, so if you said this and I missed it, I apologize.
However, there is one thing I want to emphasize that often gets overlooked in terms of influence and thickness.
Points do not just come from getting your own territory. Points come also from eliminating your opponent's territory and keeping it small. I notice in a lot of high level games, the players have actually very little territory on the board because the players limit each other so well.
A lot of people (including myself at times) want to "make points" with thickness, but this does not actually mean you need to make territory with thickness. I often see beginner players try to make thickness into a moyo and into territory, when sometimes it is better to play elsewhere and wait for the opponent to enter. In this way, we use thickness to minimize an opponent's territory rather than to increase our own territory, which has the effect of increasing our points.
The example I like to use of influence that is not thick is the 3-3 shoulder hit joseki as in http://www.lifein19x19.com/forum/viewto ... 803#p88803. That post includes a game where Sakata shows it is not thick by attacking it ferociously
$0.01 (pro): is too close to W.
$0.02 (pro): Even if we ignore and the rest of the board,
if we just look at the upper right corner, the result is already good for W --
B makes a few points in the corner, but W's immediate points (the ones directly touching W)
are already bigger than B's; moreover, W has development while B has none.
Because of B's premature ( ) jump into 3-3, W could get a good result blocking either side.
$0.03 (my feeling): the caption "thickness lost" is a misnomer.
Thanks to B's two questionable moves ( and ),
the result is both locally and globally good for White,
so if anything, it is Black that's a little "lost."
Ok, further thoughts, and explanation as to the point of the thread.
I think we need an alternative to . The purpose was to create a contrived position where thickness existed for the purpose of trying to illustrate the urgency of at least addressing thickness facing an open side. The fact that is a rubbish move probably detracts from the aim, so an alternative opening may help more - even if that constitutes some approach/pincer/3-3 joseki.
Secondly, it's an interesting point on - again, I was trying to illustrate the urgency of the area, but an overstretch may again undermine the underlying point. It might actually be useful to have a continuation for White there as to why thickness needs to be handled with care, and I'm not sure I would be able to punish very well.
Thirdly, the original purpose of the thread was to encourage making a plan based on factors on the whole board. I'm quietly hopeful it may be able to be a useful collaborative short piece eventually if good comments continue to come back, but I also don't want to get bogged down in the intricacies of subtle errors on individual variations. The biggest issue I seem to encounter from helping players between 8k and 15k is a lack of cohesive plan.
I wanted to do a few little articles like this, tightened up with community support, that could then find their way into the beginners section as helpful idea guides: Thickness/influence/territory and how they impact the important areas of the board (this one), 3rd/4th line stones in fuseki and how they open up good and bad development ideas (related but not quite the same), weak groups and how / when to attack (another tricky topic), and probably more as I think of them.
The idea isn't to give a bunch of variations apart from where contextually necessary, but to help players not feel "lost" in the opening and early midgame when decisions have to be made that don't directly relate to any book examples they might have, by providing a bunch of generalised ideas that people can come up with their own plans for.
As I once heard (second hand) from Matthew MacFadyen: "Always have a plan. Even a bad plan is much better than no plan at all".
I'm finding this useful. I still don't "get" influence or why many moves in the 15 to 30 move range are more urgent than others and the coordination with the broader board. Questions like "Is it better get my stones coordinated or better to cause disruption to my opponents coordination?" trouble me.
I often look at dan level games and see the early build of influence and thickness apparently come to nothing locally .
Although there are some glimmers. The 3rd Dan Chinese guy who is still playing me twice a week (he now has a 2 dan mate who also beats me up) occoasionally seems to start attacks from something thick which then seem to come to a focus on a distant part of the board. But it's very hard to quantify
Must admit Uberdudes 33 observation put my mind at rest. I was just responding 44 with the following moves. It's clearly a crap strategy in many circumstances..
IMO above two diag is better than below.
i prefer first over second..but i guess that is preference and can not say which is better.
below is trying to limit white's influence but can not help thinking it is loose.
Thickness has not been lost. it is still there and breathing hard on that lonely black stone that came too close.
my philosophy on thickness... if you opponent plays thick..then you have to do the same.
My ¥2: I think that there are two different senses of influence that are in common use. One is a translation of seiryoku or gaisei, which could also be translated as power or outside strength. Trading territory for influence is an example of this sense. The second sense comes from computer go (as far as I know), and means the effect that a stone or group of stones has on empty points or stones. This kind of influence could even be negative. When people talk about not understanding influence, they generally mean the first kind. Very few people really understand the second kind, IMO. Which is one reason that current computer programs have pretty well abandoned it.
As for the invasion on the 3-3, we all agree that it is bad. To the extent that I understand influence of the second kind, I think that it loses around 4 points. White gets almost a wrap-around wall, and a wrap-around wall is said to be worth more than 40 points. I used to think that that was an exaggeration, but now I think that it is about right. My (educated) guess is that the wall is worth around 30 points.
You do not neutralize the influence of that kind of wall in a single play. It is too strong, and radiates influence in more than one direction. Takagawa says that there are two kinds of walls, those that need an extension, and those that do not. That classification is fuzzy, but this wall definitely does not need an extension. It is very strong.
My feeling is to block the invasion on the right side, so that the wall faces in that direction. Because of the strong wall, I like . Any closer to the corner and White will pincer and make use of the wall to attack. The result is that the wall produces territory in the bottom right corner and influence towards the bottom side. Any questions? Then I like , driving Black towards the wall. One virtue of such a strong wall is that it hinders the development of the opponent's stones. Versus nirensei an approach from the top drives Black towards a friendly stone. Approaching from the left does not do that. Here the wall pays off by the territory that Black does not get.
And I agree with my colleagues that for Black the play is or a.
----
P. S. There is a good chance that I will give a presentation of my research into influence of the second kind at the U. S. Go Congress this summer, probably by long distance. There is no firm agreement yet.
The Adkins Principle: At some point, doesn't thinking have to go on?
— Winona Adkins
Takagawa says that there are two kinds of walls, those that need an extension, and those that do not. That classification is fuzzy, but this wall definitely does not need an extension.
I thought the distinction was you do not extend from a thick wall. Perhaps the determination of thickness is the fuzzy area, but it at least gives you a criterion for choosing whether or not to extend.
Takagawa says that there are two kinds of walls, those that need an extension, and those that do not. That classification is fuzzy, but this wall definitely does not need an extension.
I thought the distinction was you do not extend from a thick wall. Perhaps the determination of thickness is the fuzzy area, but it at least gives you a criterion for choosing whether or not to extend.
The criterion Takagawa uses is eye potential. However, in looking over pro games where walls are left without extensions, they do not always have obvious eye shape. So now I think of strength as one criterion. And thickness and strength are similar.
The Adkins Principle: At some point, doesn't thinking have to go on?
— Winona Adkins
This long knight would also be my move in this game, but the primary reason I didn't post it is because my reasoning is that it addresses the influence and strengthens the corner at the same time, making it a fairly useful dual purpose move, and making the left side rather attractive. I played the top middle hoshi to focus on the subject at hand, but this discussion about too close is fascinating. Is that because White can overconcentrate Black? Is it because White can actually invade between the top two Black stones and get a favourable result? How is White supposed to punish the too close-ness?
topazg wrote:Is it because White can actually invade between the top two Black stones and get a favourable result? How is White supposed to punish the too close-ness?
Isn't that always the problem with playing too close to thickness? I thought that if you're worried about a thick position the best course of action is to batten down the hatches and avoid giving the opponent anywhere to start a fight.
topazg wrote:Is it because White can actually invade between the top two Black stones and get a favourable result? How is White supposed to punish the too close-ness?
Isn't that always the problem with playing too close to thickness? I thought that if you're worried about a thick position the best course of action is to batten down the hatches and avoid giving the opponent anywhere to start a fight.
It's not clear to me how a continuation of the attack is supposed to happen? Black looks alive, White still has issues to deal with, and Black's building points on the left. I'm wondering if the other approach makes Black just have a suboptimal result, but I'm not clear on that either:
Play somewhere else. To me the wedge stands out. The wall will help invade the top left framework, but an invasion now would be premature. No point in letting Black make a huge framework.
The Adkins Principle: At some point, doesn't thinking have to go on?
— Winona Adkins
, still has a lot of aji at "x" through "z", whereas White' position at the top right has almost nothing for Black to use as leverage. As such, Black following immediately with 
is designed as a pre-emptive protection against "a", as Black's position at the top has solidified to the point where he can play "a" and the corresponding push / squeeze (see hide tag below for the sequence) in the hope of profiting on the left. Alternatively, because of White's thickness in the top right, an alternative could be to work on developing that side of the board, simply because Black, aware of the potential of White creating two directional thickness by bending at "b", will be wanting to approach around "c" as soon as possible. For example:
, Black tries to create a situation where an easy path for White becomes much less clear. Prior to 
is too close to W.
( 
) jump into 3-3, W could get a good result blocking either side.
.
, driving Black towards the wall. One virtue of such a strong wall is that it hinders the development of the opponent's stones. Versus nirensei an approach from the top drives Black towards a friendly stone. Approaching from the left does not do that. Here the wall pays off by the territory that Black does not get. 
- 
. With such a strong wall I want Black to invade. 
or a. 